1. Field of the Invention
This invention relates to the POU domain family of transcriptional regulators (named after three of the earliest identified members of this family: Pit-1, Oct-1 and unc-86). More specifically, it relates to a novel POU domain gene which alternatively and selectively expresses at least two forms of a protein in terminally differentiating epidermal cells and hair follicles. The first expressed form of the gene contains a region which inhibits DNA binding by itself and other transcription factors. The second form apparently serves as a transcriptional activator of, for example, cytokeratin 10 (K10) and human papilloma virus-1 (HPV-1) gene expression in skin. The gene expresses proteins which vary in sequence and length at their amino termini.
2. Description of Related Art
The POU domain family of tissue-specific transcription factors express proteins having two conserved domains: a C-terminal homeodomain of approximately 60 amino acids and an N-terminal POU specific domain of approximately 81 amino acids, which domains are connected by a variable linker region. The homeodomain is a DNA binding domain which has been found to play a central role in eukaryotic gene regulation; in general, the POU specific domain enhances the affinity of POU domain proteins for certain target DNA binding sites and may also be involved in protein-protein interactions.
The first three mammalian POU domain transcription factors (Oct-1, Oct-2 and Pit-1) were characterized and described in 1988 (see, Ingraham, et al. (1988) Cell, 55: 519-529; Bodner, et al. (1988) Cell, 55: 508-518; and, Ko et al (1988) Cell, 55: 135-144). Thereafter, a strategy was described for identifying and isolating additional members, if any, of the POU domain family using the polymerase chain reaction (PCR) and degenerate oligonucleotides representing all possible codons for two 9-amino-acid residues conserved among Oct-.1, Oct-2, Pit-1 and a C. elegans regulatory gene product, unc-86 (see, He, et al. (1989) Nature 340: 35-42). Following this technique, POU domain genes have been isolated from a variety of organisms and show diverse patterns of expression. Proteins encoded by POU domain genes include several which bind to a specific octamer DNA motif (5'-ATGCAAAT-3'), which is believed to be required for ubiquitous and tissue-specific expression of various genes. These proteins (commonly known as Oct-1 through Oct-10 in the mouse and by alternative names in humans) have generally been found to be present in embryonic and adult tissue in the mouse and humans as follows:
__________________________________________________________________________ EXPRESSION Oct Proteins Synonyms Gene Embryo Adult __________________________________________________________________________ Oct-1 Human: Oct-1 Ubiquitous Ubiquitous OTF-1; NF-A1 (1) NF-III; OBP 100 Oct-2 Human: Oct-2 Neural tube; in Lymphoid cells; OTF-2; NF-A2 (7) entire brain except nervous system; N-Oct.sub.-- telencephalon intestine; testis; kidney N-Oct-2 Human: -- Nervous system Nervous system N-Oct.sub.-- -- (Astrocytes, certain glioblastoma and neuralblastoma cell lines) MiniOct-2 -- Oct-2 Nervous system (strong Nervous system (7) expression in developing primary spermatids nasal neuroepithelium) N-Oct-3 -- -- Nervous system Nervous system (certain glioblastoma and neuroblasoma cell lines) Oct-4A NF-A3; Oct-3 Oct-4 Totipotent and Oocytes Oct-4B -- (17) pluripotent stem cells of Oct-5 -- the pregastrulation embryo, embryonic ectoderm, primodial germ cells; testis, ovary Oct-6 Rat: Tst-1; SCIP Oct-6 Blastocyst; ES and EC Nervous system (4) cells; brain testis Oct-7 Human: N-Oct-4 -- Nervous system Nervous system Oct-8 Human: -- Nervous system Nervous system Oct-9 N-Oct5a (astrocytes) Oct-10 N-Oct5b __________________________________________________________________________
Although the above list should not be regarded as exhaustive, it serves to demonstrate the predominance of many members of this gene family in orchestrating precise temporal and spatial expression of genes during development of the neural system (particularly the forebrain), endocrine system and, for Oct-4 and Oct-6, embryonic and adult reproductive tissues.
Consistent with this expected distribution, part of the POU domain of an eleventh Oct gene was identified as being present (but not functional) in mouse testis by PCR and a proposed partial sequence of the gene published in 1990 (Goldsborough, et al. (1990) Nuc. Acids. Res. 18: 1634). Although not prior art, it is notable that more recent (1993) data published by these same authors led them to the conclusion that this eleventh gene (referred to as "Oct-11") is expressed in a tissue-specific manner during mouse embryogenesis and in adult thymus and testis tissue as well as myeloma cell line P3/NS-1/1-Ag4.1. However, the authors also concluded that only one form of the gene was present in rats and humans and only one active form was present in the mouse genome.
The results reported by Goldsborough, et al. did not predict or suggest either that quantity of expression of the "Oct-11" gene or its potential function. Their results regarding the distribution of the Oct-11 transcript (as reported in the 1990 Nuc. Acids. Res. article) were necessarily limited by the sensitive, nonquantitative nature of the PCR methods used. Specifically, as known in the art, the use of PCR to identify genes suggested in the 1989 He, et al. Nature article will amplify DNA ,expressed from genes present in tissues even in minute quantities, regardless of whether a function is served by the gene in the source tissue. Therefore, while helpful in identifying the existence of potentially new genes, the PCR approach used by Goldsborough, et al. does not reveal patterns of expression or function for identified genes (see, e.g., Pfaffle, et al. (1992) Science 257: 1118-1121. reporting the presence of Pit-1 [which is only functional in the pituitary] in lymphoid cells).
In contrast, based on data developed prior to the 1993 publication by Goldsborough, et al., and contrary to their conclusions therein, not only are two functional forms of an eleventh Oct gene expressed in both rats and humans, the genes are not limited in their distribution to testis and thymus tissue or the P3/NS-1/1Ag4.1 cell line. Instead, alternative full-length forms of these genes (hereafter respectively referred to as "Skn-1a" and "Skn-1i") are selectively expressed in terminally differentiating epidermal cells and hair follicles. In this context, distinct regions of Skn-1a/i play functional roles in regulating skin development.
For perspective, it should be considered that while skin is the largest organ in mature mammals, cell-specific transcription factors involved in epidermal cell maturation have been previously unknown (although several factors of wide distribution are known to be expressed in skin, including AP2, retionic acid receptor .gamma. and retinoid X receptor .alpha.). Therefore, an understanding of the role of Skn-1i/a in epidermal cell development may be facilitated by a brief illustration of the pattern of this development in the rat.
Epidermal development in the rat does not begin until embryonic days (e) 15 to 16. Before this stage, the primordium of epidermis consists of a bilayer of cells, a superficial layer referred to as periderm that is later shed and a basal layer. On .theta. 16, the basal cells begin to proliferate, generating a stratified epithelium in which a characteristic subset of genes (such as keratins) are differentially regulated within each layer. Most of the suprabasal epidermal cells are post-mitotic, and eventually undergo programmed cell death, generating a superficial layer of dead cells (cornified epithelium) that appears on .theta. 18. This pattern of development, in which cells migrate to the surface during their differentiation only to undergo apoptosis, is continuously repeated in the adult, where the process is regulated by retinoic acid.
Skn-1i and Skn-1a's roles in this process appear, at least in part, to respectively be in the specific inhibition of DNA binding by transcription factors (particularly the ubiquitous Oct-1 protein) and in the control of this inhibitory activity by Skn-1a. Ultimately, controlled expression these genes may not only lead to further insight into the regulation of skin development but may also assist in the clinical control of skin cell proliferation for use in, for example, regenerating skin and terminating metastasis of tumor cells.